surface chemistry of fluorochemicals

0 NRL Report 6324

Surface Chemistry of Fluorochemicals

N. L JARVIS AND W. A. ZISMAN

Surface Chemittry Branch

Chemistry Division C L E A R I M G H i ti

"O,, rEDERAL SCIENI I *P' ",)TEGHNICAL INFOR1M• f

I ,. -

October 21, 1965

DDC

DEC16 1965J

I ISIA B

U.S. NAVAL RESEARCH LABORATORYWashington, D.C.

Best Available Copy

7

CONTENTS

Page

Abstract iiProblem Status iiAuthorization ii

INTRODUCTION 1

LIQUID SPREADING AND ADHESION ON LOW-ENEiiuxSOLID SURFACES 1

EXTENT OF FLUORINATION REQUIRED F0 ..OW 8

MONOLAYERS OF HALOGENATED ORGANIC COMPOUNDSADSORBED ON SOLID SURFACES 9

THE REVERSIBLE WORK OF LIQUID ADHESION TO SOLIDS 10

SURFACE ACTIVITY OF FLUOROCHEMICALS IN AQULOUSSOLUTIONS 11

INSOLUBLE FLUOROCHEMICAL MONOLAYERS AT THELIQUID/GAS INTERFACE 19

SURFACE ACTIVITY OF FLUOROCHEMICALS IN NON-AQUEOUS SOLUTIONS 20

ADHESIVE AND ABHESIVE BEHAVIOR OF FLUOROCHEMI-CALS 23

CONCLUSIONS 27

RE FERENCES 29

ABSTRACT

Research on the surface chemistry of fluorochemicals, including the related deveiop-ment of new fluorine-containing compounds, polymers, and surface treatments, has pro-gressed to the stage where a review of the subject is appropriate. A report has been pre-pared in the hope that it will serve not only as a record of progress, but will also clarifymuch of the fundamental information that is available, and will serve as a guide for futureinvestigations and applications.

With the advent of the highly fluorinated organic compounds, it has been possible toprepare surfaces having extremely low free surface energies, much lower than had pre-viously been possible with hydrocarbon derivatives. Because of their low surface energies,fluorinated solids have the most nonwettable and nonadhesive surfaces known. It hasbeen proved that only the outermost molecules of the surface of a solid need be highlyfluorinated in order to achieve such low surface energies. In fact the most nonwettingsurfaces studied to date were those covered with an adsorbed monolayer of closely packedperfluoro acids, with the perfluoromethyl (-CF 3 ) groups outermost.

It has likewise been found that surface-active agents containing perfluorocarbon ter-minal groups will lower the surface tension of water far below the values obtainable withhydrocarbon-type wetting agents. Certain fluorinated hydrocarbon derivatives have alsobeen synthesized that show very high surface activity in a wide variety of organic liquids,the extent of surface activity being dependent upon the organophilic-organophobic balancein the fluorine-containing amphipathic molecule.

It is predicted that fluorine-containing monomers and fluorocarbon-substitutedsurface-active materials will find increasing application in polymers and plastics, asadditives for liquids, and as protective coatings on tsolids.

PROBLEM STATUS

This is an interim report; work on this proble-n is continuing.

AUTHORIZATION

NRL Problem C02-o10Project RR 001-01-43-4751

Manuscript Submitted July 13, 1965

ii

SURFACE CHEMISTRY OF FLUOROCHEMICALS

INTRODUCTION

Since the 1940's the advent of fluorinated organic chemicals has had a major impacton science and technology. Certain fluorochemicals prepared just before or duringWorld War II have made it possible to observe greater changes in surface energy, wetta-bility, and surface activity than previously had been possible. Research on surface prop-erties of organic fluorochemicals has since revealed a host of new applications as well asseveral new areas to be explored. Another important result is that today we have a sub-stantial body of knowledge to teach us much about the relation between the chemical struc-ture and surface properties of liquids and solids.

Why do some fluorochemicals have such extreme surface properties? How doeseach fluorine atom contribute to the resulting surface activity of the molecule, and whatstructural rules are there to guide one in designing surface-active fluorochemicals?These questions can be answered in terms of what we have learned about the surfaceproperties of these fluorochemicals. The most important properties are: (a) the equilib-rium contact angle of almost any liquid on a solid surface covered with perfluoromethylor perfluoromethylene groups is much greater than on a solid surface covered with methylor methylene groups; (b) aliphatic fluorinated polar-nonpolar compounds (such as thecarboxylic acids, alcohols, and salts) are much more surface active in water than eithertheir hydrocarbon analogs or any other class of compounds; and (c) fluorocarbon deriva-tives of the organophobic-orgi"oiphilic type manifest uniquely high surface activity innearly all nonaqueous liquids. Each of these topics deserves discussion here.

LIQUID S. READING AND ADHESION ON LOW-ENERGY SOLID SURFACES

Provided the surface of a solid fluorocarbon polymer is sufficiently smooth and clean,the equilibrium contact angle ti of a sessile drop of any pure liquid on the horizontal sur-face (Fig. 1) is a reproducible and indicative property (1). A plot of cos 0 vs the surfacetension Y,,. of the liquid is shown in Fig. 2 for the homologous family of n-alkanes onpolytetrafluoroethylene. The graphical intercept with the straight line cos o = I is de-fined as the critical surface tension Y, of wetting. A similar plot using any other homol-ogous series of liquids results in practically the same value of -Y. If a similar chart isplotted for a variety of liquids without regard to their structural homology (Fig. 3), thegraphical points are nearly all distributed in a narrow rectilinear band rather than astraight line. The lower boundary of this region still defines an intercept at cos o, = Iwhich is about the same as that given in Fig. 2 and is taken as -Y, for that system.

VAPOR

Fig. I - Schematic dia-gram of finite contactangle formed by a sessiledrop resting on a solidsurfaceSOI

2 NAVAL RESEARCH LABORATORY

10 Graphical points for the more polar liquids,0 5having surface tensions above 40 dynes/cm, lie

on a concave-up curve deviating more from theextended straight line the greater yLt. The

.9 o straight-line region is due to the molecules ofthe low-surface-tensiou liquids being attracted

S0 . to the solid surface principally by the London"dispersion" forces. The deviation from line-

3.8 arity occurs because of the additional forces ofattraction which are possible with the morepolar, higher-surface-tension liquids, such as

0 the formation of weak hydrogen bonds betweenthe fluorine atoms in the surface of polytetra-

.1 •fluoroethylene and the hydrogen atoms of thehydrogen-donating liquids. Figure 3 exempli-fies the results of many similar studies of other

18 20 22 24 26 28 types of polymeric solids.SURFACE TENSION AT 20C (DYNES/CM)

A most revealing study was the effect onFig. 2 - Wettabilityof poly- -y of the progressive fluorination (or chlorina-tetrafluoroethylene by n- tion) of linear polyethylene. Table 1 comparesalkanes values of Y, for the various halogenated

*o • n-AI.ANES ISOo oTHER HYOOCARO.NO

08- U ESTERS AND ETHERS0.8 y HALOCARBONS AND

NALOHYOROCARSONS -48£ MISCELLANEOUS LIQUIDS

06-

t 0

o° O n-ALAS !,• I0s

.0 L TISS

ig. ~ ~ OHE HYRCABN -Wetb3t0o oyeralooty

K IPOLVTTRANUOROEHTLEE

\

0 10 20 30 40 OC60 ?O 0SUEC ESO L20 C) DYNEOS L'O

lene by water and a variety of organic liq-uids

polyethylenes and relates them to the polymeric structure and the degree of halogenationof each solid surface. It is assumed that the surface composition of each polymericsolid is statistically the equivalent of that of an individual polymer molecule stretchedout in a horizontal configuration at the surface. Thus, the surfaces of linear, unbranchedpolyethylene and polytetrafluoroethylene are presumed to be comprised essentially of-CH 2 - groups and of -CF 2 - groups, respectively, poly(vinyl fluoride) to have a surfaceatomic composition of 25 percent vuwalent fluorine and 75 percent covalent hydrogen, etc.One can then plot the ~,, data of Table 2 against the estimated atom percent halogenationof polyethylene as shown in Fig. 4. In striking contrast are the descending straight-line

- C R B N A N D - - ,,e e.

NAVAL RESEARCH LABORATORY 3

Table 1Critical Surface Tension (-y) and Constitution

of Halogenated Polyethylenes

P Critical SurfacePolymer Structural Formula Tension (dynes/cm)

H CI H C1 H C1I I I I I I

Poly(vinylidene -C-C -C-C -C-C 40chloride)

H C1 H Cl H Cl

H C1 H Cl H C1Poly(vinyl chloride) -C-C -C-C -C-C 39

I I I I I I

HH HH HH

HH HH HHI I I II I

Polyethylene -C-C C-C- -C-C 31II I t I

HH HH HH

HF HF HFPoly(vinyl fluoride) -C -C, -C- 28

I I I I I I

HH HH HHHF HF HF

Poly(vinylidene -C-C C-C- -C-C 25fluoride) I I II I I

H F H F H F

FF FF FFI I I I I I

Polytrifluoroethylene -C-C -C-C -C-C 22i !I I I I

HF HF HF

FF FF FF

Polytetrafluoroethylene -C-C -C- -C-C 18I I I I I I

(Teflon) FF F F F F

behavior of the fluorohydrocarbons and the ascending asymptotic curve of the chlorohydro-carbons. As will be shown, cos o is a measure of the adhesion of a given liquid to thesolid; hence Fig. 4 states that surface fluorination decreases and surface chlorination in-creases the adhesion of each liquid to the solid. Such a major difference in behavior be-tween covalent fluorine and chlorine is of much surface-chemical importance, but it is notunique to the family of halogens in view of many other anomalies in behavior of the firstelement of a periodic group.

The extrapolation of the upper graph of Fig.ý 4 to 100 percent chlorination predicts avalue of 40 dynes/cm for a solid surface coated with close-packed covalent chlorine atoms.Fully chlorine-substituted polyethylene is not available, and presumably it cannot be pre-pared. Since the chlorine atom is much larger than hydrogen, steric hindrance preventsthe complete replacement of all the hydrogen atoms. However, perchloropenta-2,4-dienolcacid (Fig. 5) has been adsorbed on polished glass and platinum to form condensed mono-layers arranged with each carboxylic acid group in contact with the solid surface and

--. -. w...


Table 2Critical Surface Tension (ye) of Low-Energy Surfaces

Surface Constitution T (dynes/cm at 20*C)

Fluorocarbon Surfaces

-CF 3 6

-CF 2H 15-CF 3 and -CF 2 - 17

-CF 2-CF 2 - 18-CF 2 -CFH- 22

-CF 2 -CH 2 - 25

-CFH-CH 2 - 28

Hydrocarbon Surfaces

-CH 3 (crystal) 20-22

-CH 3 (monolayer) 22-24

-CH 2-CH 2- 31

-CH- (phenyl ring edge) 35

Chlorocarbon Surfaces

-CCIH-CH 2- 39

-CCI 2-CH 2 - 40

=CC12 43

Nitrated Hydrocarbon Surfaces

-CH 2ONO 2 (crystal) [ 110] 40

-C(NO2)3 (monolayer) 42

-CH 2ONO 2 (crystal) [101] 45

with the two covalent chlorine atoms of the e,-carbon directed outermost. The v0 forthis coated solid was 41 dynes/cm, in good agreement with the extrapolated value. Thisremarkable result exemplifies the general rule that the wettability of an organic surfaceis determined solely by the composition and orientation of the outermost P, oms in thesurface of the solid and is independent of the nature of the underlying material (1). Thisrule appears correct so long as materials are avoided which contain trapped ioniccharges or large unbalanced electrostatic dipoles in the organic surface.

Numerous other polar-nonpolar compounds besides perchloropentadienoic acid havebeen adsorbed as close-packed monolayers on metals, glass, and other smooth high-energy solid surfaces, and the cos o vs I,,v graphs of the resulting coated materials havebeen obtained. Such compounds have included homologous series of fatty acids, amines,ami"' ., and alcohols, as well as a variety of halogenated derivatives. Wetting proper-ties • single crystals of the paraffin hexatriacontane (C36 H 74 ) also have been investi-gated. In this crystal the outermost composition of the freshly cleaved surface is


comprised of oriented methyl groups in 0 60

close packing. It is instructive to com- bpare the value of y,. = 22 dynes/cm of sosuch a surface with that of 24 dynes/cmof a close-packed monolayer of stearic BY CHLORINEacid adsorbed on platinum, chromium, • 40 - -- .....iron, aluminum, zinc, nickel, beryllium,or soda-lime glass. The decrease in -ý

of 2 dynes/cm in going from the adsorbed Z

monolayers to hexatriacontane is rea-sonable, since it results from the slightly X 20closer packing of the methyl groups inthe latter material.

.J 4

In 1947 the value of ),, for polyethyl- -

ene was compared by Zisman with thatfor polytetrafluoruethylene, and the value 0 0 215 5o 05 ,oof .jfor a surface of close-packed methyl HYDROGEN REPLACED (ATOM PERCENT)

groups with that for a surface of close- Fig. 4 - Thc effect of progressive hal-packed perfluoromethyl groups. The fact - ogen substitution on the wettability ofthat ,. of close-packed -CH 2- groups is polyethylene-type surfaces31 dynes/cm and that of -CH 3 groups inhexatriacontane was only 22 dynes/cmled him to estimate that .y for condensed CI ci-CF 3 groups must be much less than the cvalue of 18.5 dynes/cm for a surface ofclose-packed -CF 2 - groups. Hence, just cF C5

as the surface of close-packed methyl Fig. 5 - Structure of perchlo- Iropentadienoic acid C1-Cgroups in a condensed monolayer of a 0fatty acid or fatty primary amine has the c. c1lo-vest surface energy of any aliphatic I

hydrocarbon film, so should the surface 0 ;;" eofof close-packed perfluoromethyl groupshave the lowest surface energy of anyfluorochemical film. This argument ledZisman and co-workers (2-5) to investigate the wettability properties of a condensedmonolayer of each of the perfluoro fatty acids as soon as it became available by theSimons electrochemical process (6). Acids prepared by the telomerization method (7,8)were also studied; these have a terminal -CF 2 H group. Some of the results of these andrelated studies are compared In Fig. 6. As predicted, a condensed monolayer of any ad-sorbed perfluoroalkanoic acid had a much lower Y, than that of any hydrocarbon surfaceor of polytetrafluoroethylene; indeed, for the most condensed film yet investigated (per-fluorolauric acid) 1y was about 6 dynes/cm, the lowest value of ), reported. As will beshown, this result Is fundamental in explaining the surface chemistry of fluorocarbonsand fluorohydrocarbons, and therefore it is important in surface chemistry.

The significance of such low values of Y,. is clearly revealed in Fig. 7. The equilib-rium contact angles at 20 0C of a selected group of pure liquids were measured onsmooth solid surfaces coated with a condensed monolayer of a fluorinated acid. Suchlarge contact angles by organic liquids are spectacular! Here cos o is plotted against thenumber of fluorinated carbon atoms per molecule of the adsorbed acid monolayers. Theseresults have suggested a variety of new areas of application: unquestionably, they pointedthe way to the now well-known surface treatments for oil, water, and stain-resistanttextiles (9-12) and leather (13,14), to new promoters of dropwise condensation for distil-lation processes, and to new barrier materials to prevent liquid spreading on ball bear-ings, watches, and other devices (15-18).


"I n

% '% 30

08- '. NST'EAR.tC ACID MONOLAYER

NEXATRIACONTANk 45POLY TE TRAFLUOROLE THYLENE

V)L,j

S-60LL 0

04PEFRFLUOROBUTYRIC ACID MONOLAIrRPERFLUOROVALERIC ACID MONOLAYERPERFLUOROCAPROIC ACID MONOLAYERPERFLUOROCAPRYL IC ACID MONOLAYER_ 75PERP tJOROCAPRIC ACID MONOLAYER

02: PERFLUOROLAU!IC ACiO MONOLAYER

0- 1 1 90

20 24 28 32 36 40

SURFACE TENSION (20 0C)DYNES/CM

Fig. 6 - Comparison of vs cos -, relation-ships for a series of n-alkanes on several differ-ent low energy surfaceb

IT

08- ýPERFLUOROALKANE 45

0e-660

04-

EO IS(2-ETHYLHEXYL) 7 "z02- SEBACATE

TERT-BUTYL

0 NAPHTHALENE,ETRACH LORO-

SI pHENYL-02 -METHYLENE

IODIDE -05

"-04 .1 1- I I I0 2 4 6 8 K 12 K 6

Nz TOTAL CARBON ATOMS/ACIO MOLECULE

Fig. 7 - the effect of homology in the fluor-inated acid monolayers on the contact anglesobserved for some miscellaneous fluids


Widespread occurrence of the rectilinear relationship between cos t, and Y, r in thelarge body of experimental data, and the fact that generally these graphs do not cross,make it possible to use ),, to characterize and compare the wettability of a wide varietyof low-energy surfaces (1). Because wettability, and hence Y,, depends solely upon thenature and packing of the outermost atoms in the solid surface, the results of many in-vestigations of the wettabilities of well-defined, low-energy, solid surfaces can be sum-marized in the very suggestive form shown in Table 2. In the first column is given theatomic constitution of the organic radicals in the solid surface, and in the second columnare the values of -- arranged in increasing order. The data have been grouped undersubheadings emphasizing the atomic components of primary interest: fluorohydrocar-bons and fluorocarbons, hydrocarbons, chlorocarbons and chlorohydrocarbons, and ni-trated hydrocarbons. The implications of Table 2 have been discussed more fully else-where (1,19), but it is essential to point out here that the first group-the fluorochemi-cals-have surfaces of lowest -y,. and lowest surface energy. Note that the lowest valueof •,. is that of close-packed perfluoromethyl groups. Replacement of a single fluorineatom by a hydrogen atom in a terminal -CF 3 group has the effect of more than doubling),,. For convenience of reference, values of -y, for smooth, clean surfaces of some im-portant linear organic polymeric solids, are arranged in increasing order in Table 3.

Table 3Critical Surface Tension (vt) of Various Polymeric Solids

Polymeric Solid J (dynes/cm at 20*C)

Polymethacrylic Ester of 1H,IH- 10.6

pentadecafluoro-1 -octanol

Polyhexafluoropropylene 16.2

Polytetrafluoroethylene 18.5

Polytrifluoroethylene 22

Poly(vinylidene fluoride) 25

Poly(vinyl fluoride) 28

Polyethylene 31

Polychlorotrifluoroethylene 31

Polystyrene 33

Poly(vinyl alcohol) 37

Poly(vinyl chloride) 40

Poly(vinylidene chloride) 40

Polyhexamethylene adlpamide 46

(Nylon 6,6).

NAVAL RESEARCH LABORATORY

EXTENT OF FLUORINATION REQUIRED FOR LOW •

What factors determine the minimum number of fluorine atoms needed per organicmolecule in order to obtain a very low value of y,? This question is important, sincefully fluorinated compounds including polymers are difficult to prepare and hence areexpensive. The only answers we can give with confidence at present relate to the .,values ! .'- unbranched aliphatic structures having different extents of fluorination. Theseanswers have resulted from studies of condensed monolayers of , -trifluorooctadecanoicacid and . - trifluorooctadecyl amine, both compounds prepared by Gavlin and Maguire(20). Condensed monolayers of a series of progressively fluorinated fatty acids preparedby Brace (21) were also studied.

These organic structures are exemplified by the diagram in Fig. 8 of a carboxylicacid comprising a perfluorinated aliphatic terminal segment attached to an unfluorinatedfatty acid segment. The terminal perfluorinated segments varied from one to ten carbon

atoms; the unfluorinated segments were the omega-substituted undecanoic and heptadecanoic acids. The

F total number of carbon atoms per molecule in the-- acids studied varied from 17 to 23. Each segmented

F-C -F acid was studied as an adsorbed condensed monolayerF-C-F on polished pure chromium and nickel, and the re-F-C-F suiting measurements of , are plotted in Fig. 9F-C--F against the number (X) of fluorinated carbon atoms.F-I-F The dotted curve is a corresponding plot for theF- -F homolgous series of perfluoro-n-alkanoic acids ad-

sorbed as condensed films on the same solid (22). ItF-C- F will be seen that complete fluorination of-the terminalF-C - F methyl group caused only a small decrease in .. NotF-C-F Fig. 8 - Structure of until all the hydrogens on the last seven or eight car-H-C-H I 1- (perfluorodecyl)H-C,-H I ndper rodc acid bon atoms in the aliphatic chain (Fig. 9) have been re-H-,-H placed by fluorine did ., decrease to the value ob-H-C -H tained with the perfluoro fatty acid monolayer containing

the same amount of fluorine. Therefore, the gainsH-6-H In using these segmented, partially fluorinated fattyH-C-H acids are: (a) they avoid the high acidity of the per-

fluoro carboxylic acids, and (b) they allow one to ob-

H- -Ht tain a monolayer coating which has much less vola-H- -H tility than the analogous fatty acid containing the same

amount of fluorine.o.Ho

The difficulties in achieving the desired lowcritical surface tension by merely fluorinating one ora few of the terminal aliphatic carbon atoms appear

to contradict the previous statement about the constitutive law of wetting. The difficultyarises from the presence of uncompensated electrostatic dipoles in the adsorbed partiallyfluorinated fatty a'!ids and their effe -t in decreasing adlineation of the molecules in thecondensed monolayer. It is well known that the electrostatic dipole moment contributionsof the various carbon-hydrogen and other bonds can often be added up vectorially to ac-count (within a good approximation) for the total dipole moment of the molecule. In thehigher n-alkane derivatives, the total dipole moment of the chain of -CH 2 - groups is zero;hence, in the case of the fatty acids the only unbalanced dipoles arise from the contribu-tions of the terminal -CH 3 and -COOH groups. Whereas the terminal methyl group makesa dipole contribution estimated to be 0.3 to 0.4 Debye unit (23), the terminal perfluoro-methyl dipole is about 1.9 Debyes (24). When the molecules are adsorbed as a condensedoriented monolayer, these terminal dipoles repel each other even though their momentsare altered by laterally induced polarization; however, the dipole moment of the -CF,,group remains at least 0.5 Debye in the direction normal to the interface (25). The large


2 -4

0 F(CF,),XCOOH (FROM SOLUTION)

o0 & F(CFt)x(CH)I 6 C0O0H (9ROM SOLUTrX

LAz2 A FtCF2Ix(CN• 1*COOH (FROM M(LT2

2"' 6S 0l N(CH 2),,OOH (PROM SOLUTION)

_JI--

W 4-In-o--jI-•

SI I I I I Io 2 4 6 0 10 2 14 1 * I

XINUMSER OF FLUORINATED CARBON ATOMS PER MOLECULE

Fig. 9 - The effect of, fluorination of the ad-sorbed acid monolayer on its critical surfacetension of wetting by n-alkanes

dipole present in the outermost surface of the partially fluorinated monolayer makes acontribution to the adhesion of the liquid which is in addition to that arising from theLondon "dispersion" forces. As the molecule is progressively fluorinated, the dipoleof the -CF 3 terminal may be compensated, in part, by the contributions of the increasingnumber of -CF 2 - groups attached to it; however, there remains an unbalanced dipole atthe junction of the fluorinated segments of the aliphatic chain. This junction dipole isfurther separated from the adhering liquid as progressive fluorination continues; itscontribution to liquid adhesion becomes negligible when there are several or more fluo-rinated atoms in the chain. The repulsion between these uncompensated dipoles, how-ever, will also cause the fluorinated segments of neighboring molecules to separatesomewhat. This effect appears significant only when X = 1 or 2. The disruptior of chainadlineation becomes minor as the number of CF 2 groups, and the contribution of the at-tractive "dispersion" force, increases. We conclude that the monolayer is a two-dimensional liquidat low values of X. At higher values it becomes a plastic solid film. Ap-parently the same packing as a perfluorooctanoic acid monolayer develops when X > 7.The need for overcoming the repulsion effect explains the need for a long fluorinatedaliphatic segment in the acid molecule.

MONOLAYERS OF HALOGENATED ORGANIC COMPOUNDSADSORBED ON SOLID SURFACES

Much has been learned in the past 20 years about the adsorption of organic mono-layers on solid surfaces and the properties of the resulting coated surfaces. Especiallypertinent to this review is the result of the NRL investigations of the adsorbed mono-layers of organic halocarbon derivatives - particularly those containing fluorine. Ex-amples of the importance of the properties of such adsorbed monolayers have alreadybeen given in discussing surfaces covered with perchloropentadienoic acid (Fig. 5), aswell as those covered with oriented close-packed perfluoro-alkanoic acids. In these in-vestigations several methods were used to prepare the nionolayers and observe theirproperties. These were the retraction method, the autophobic liquid technique (a specialcase of the retraction method), and a modification of the rubbing-down method of SirWilliam Hardy (26). The first two were developed and investigated fully by Zisman (1),


and the third niethod was recently studied and its limitations discussed, by Levine andZisman (27,28) and I3ewig and Zisrnan (29).

The retraction method, which is a direct consequence of the constitutive law of wet-ting, is a process in which a clean, high-energy plane surface is first contacted with aliquid containing adsorbable, polar-nonpolar molecules in solution. Adsorption of theamphipathic molecules occurs at the solid-liquid interlace, and the solid is covered byan oriented monolayer which converts it to a low-energy surface. When the plane of thecoated solid is held approximately vertically during its withdrawal from the solution, thesolution exhibits an appreciable contact angle and then peels back (or retracts), leavingthe withdrawn surface of the solid dry but still coated with the adsorbed oriented mono-laye-r. The explanation for this behavior is simple (1,19). The monolayer-coated surfacecan be removed from the solution unwetted only when withdrawn slowly, and then onlywhen the contact angle of the liquid with the solid is greater than zero. However, 0only for those liquids having surface tensions .,., greater than the critical surface ten-sion of the monolayer-coated solid suriace; therefore, the necessary condition for re-traction is simply that ., . Smooth, plane solids covered with adsorbed monolayerscan thus be isolated from solutions having surface tensions greater than the values ofof the low-energy surfaces created by the solution-adsorption process.

Monolayers of a wide variety of organic polar-nonpolar compounds have been pre-pared on solid surfaces by the preceding methods, and many of their physical and chemi-cal properties have been investigated without the experimental difficulties encounteredwhenever the solution phase is present. Shafrin and Zisman (30) have reviewed the earlyNRL investigations of a wide variety of polar-nonpolar compounds (alipnatic, cyclic-aliphatic, and aromatic) adsorbed on platinum from aqueous solutions by the retractionmethod. Monolayers adsorbed from nonaqueous solutions or from the melt have beenwidely investigated (1,31), especially the families of fatty acids, alcohols, primaryamines, and amides. Many classes of halogenated organic compounds were studies, includingthe perfluoroalkanoic acids (3,27,28,32,33), the telomer acids, HF 2C(CF2)xCOOH (4,27,28,32,33), perchloropentadienoic acid (27,28,34), -trifluorostearic acid and -trifluoro-stearyl amine (27,28,33,35), the progressively fluorinated alkanoic acids (21,22), thedirner and trimer acids (27,28) of the series Cl(CF 2 CFCl)xCF 2 "COOH, and a miscellanyof other halogenated acids (27,28) containing F, Cl, or Br.

Properties most thoroughly investigated included contact angles and ,,, surface po-tentials, and coefficients of friction. Additional properties are being investigated at NRLand many other laboratories. An interesting radioactive tracer study of perfluorooctanoicacid was made by Shepard and Ryan (36), and a miscellany of friction studies were madeby Bowden, Tabor, and co-workers (37).

These investigations of the properties of retracted monolayers adsorbed on solidshave been basic contributions to surface chemistry, and they have already thrown muchlight on a variety of applied problems in wetting, adhesion, friction and wear, electricalproperties of surfaces, and corrosion inhibition. For example, Table 4 on abhesive agentsis really a byproduct of these studies.

THE REVERSIBLE WORK OF LIQUID ADHESION TO SOLIDS

The reversible work on adhesion it I of a liquid of surface tension :, v to a solid canbe computed from the equilibrium contact angle and the free energy of immersion f.of the solid in the liquid vapor by the well-known relation of Bangham and Razouk (38,39):


Table 4Critical Surface Tension (. ) of Surfaces Coated withl Abhesives

Coating Material (dynes/cm(All Condensed)J at 20'C)

Polymethylsiloxane film 24

Fatty acid monolayer 24

Pulvtetcrafluoroethylene film 18

HC F (C F,),, COOH monolayer 15

Polymethacrylic ester ofIH,1H pentadecafluoro-1-octanol 10

Perfluorolauric acid monolayer 6

At present there are no data available on I L for solids of low surface energy. How-ever, if is not too close to zero, it canbe, assumed that f,,, will be negligible (1); undersuch circumstances the reversible work of adhesion is given approximately by it , where

II I. '/. I ( I - ,,- ;t' . (2 )

This approximation is equivalent to assuming that the amount of adsorbed vapor of theliquid on the low-energy solid surface is much less than a condensed monolayer ý,t ordi-nary temperatures. Bewig and Zisman (29) have given surface-potential data to supportthe correctness of this assumptior. for polytetrafluoroethylene surfaces; Martinet (40)and Graham (41) recently have published adsorption isotherms for the same solid whichlead to essentially the same conclusion, so long as extremely low temperatures are notinvolved.

As:suming that the approximation given by Eq. (2) is correct, and recognizing fromFig. I that cos is linearly related to ./A by

I-,.. I , (3)

where . is an expcrimental constant, it follows that it1 . is the secopd-degree parabolicfunction of ", & given in Eq. (4)

...... • ' ," - '(4)

In Figs. 10 and 11 are plotted some of the data for i . vs '.,, for various kinds of fluori-nated surfaces, and in Table 5 are given the maximum values of it . read from suchcharts. Note that the range from highest to lowest .. is about 6 to 1 and that the corre-sponding ratio of (it ,) is about 3 to 1.

SURFACE ACTIVITY OF FLUOROCHEMICALS IN AQUEOUS SOLUTIONS

Gibbs' classic work (42) proved thermodynamically that any compound in solutionwhich manifests the ability to depress the surface tension (or free surface energy) of thesolution must at equilibrium be adsorbed at the-gas/solution interface. Until about 1950the surface-active agents in use were hydrocarbon derivatives consisting of one or morehydrophobic groups 'always hydrocarbon or hydrocarbon derivatives) to which were


11O* - POLYVINYL FLUORIDEo0- POLYVINYLIDENE FLUORIDEA- POLY TRIFLUOROE THYLENE

100 - 0- POLYETHYLENEX POLYTETRAFLUOROETHYLIENEO POLYHEXAFLUOROPROPYLENE7- POLYMETHACRYLIC ESTER

90- OF F(CF 2)7 CH.2OH

S80 0

U

0 70

60

"20 3O 40 50 60 70 80SURFACE TENSION (DYNES/CM AT 200C)

Fig. 10 - Effect of liquid surface tension on It1. forlow surface energy polymeric solids

attached one or more hydrophilic polar groups. A zone chart (Fig. 12) by Fischer andGans (43) summarizes the surface-tension-depressant behavior of a representative groupof such compounds. For convenience, we will refer to such compounds as "conventional"wetting agents. The chart makes it evident that the lowest surface tension of water at-tainable with any such compound at 20'C is between 25 and 27 dynes/cm.

What determines this minimum surface tension? The answer is most instructive,because the surface tension of the solvent attains its minimum value when the solutemolecules are in closest possible packing with the principal hydrophobic chains directedout of the water into the gas phase. If the hydrophobic portion of the solute molecule isan unbranched aliphatic chain, if the molecules are not ionized, and if the polar groupsare of smaller cross section than the aliphatic chain, the maximum surface-tensionlowering would be expected when adlineated paraffin chains of the solute monolayer reachclosest packing. For example, when a lauryl alcohol monolayer is adsorbed on the sur-face of water as a fully oriented film in closest possible packing, the surface-tensionminimum will be attained. It was pointed out earlier by Jarvis and Zisman (44) that whenthe outermost portion of the oriented monolayer is comprised of n-alkyl chains whosemethyl groups approach closest packing, the value.of ). of that surface should be at leastas low as 24 dynes/cm at 20'C. They suggested that the lowest surface tension obtain-able with any single, pure conventional surface-active agent In any solvent at 20 0C wouldalso approach 24 dynes/cm. The fact that conventional wetting agents do not lower the


100

90

80

2 70U

'2A

60

40 --

0 13_ •3 CF0

40- X -CF 3

30 I I20 30 40 50 60 70 80 90

SURFACE TENSION (DYNES/CM AT 200C)

Fig. 11 - Effect of liquid surface tension on 1.i . fora solid coated with a condensed monolayer

surface tension of water below 25 to 27 dynes/cm would then be understandable, becausethese agents must be soluble in water to some extent for the use intended. The problemof attaining closest possible packing of the adsorbed film is, therefore, not easily an-swered with such a material without creating conditions leading to precipitation of thesolute.

One can obtain lower surface tensions, however, by using a slightly insolublesurface-active wetting agent like lauryl alcohol in combination with a material whichhelps solubilize it in water. Such a solubilizing agent would be a surface-active com-pound that forms micelles. Such synergistic surface-active systems have been studiedfor many years by various investigators, even since 1944, when Miles and Shedlovsky(45) first reported the synergistic effect of small concentrations of lauryl alcohol addedto sodium lauryl sulphate to lower the minimum surface tension obtainable from about 38dynes/cm to about 29 dynes/cm at 25°C. Burcik and Newman (46) further investigatedthis combination of agents and showed that with 7 percent lauryl alcohol present, the sur-face tension of sodium lauryl sulphate could be lowered to 22.1 dynes/cm at 25 0 C. Harva(47) showed that the minimum attainable surface tension of 36 dynes/cm of pure sodiumlaurate could be reduced to about 23 dynes/cm by the addition of 2 X 10-4 molal 1-decanol.A similar minimum value could be obtained by substituting 1-octanol, 1-heptanol, or 1-hexanol for the 1-decanol simply by using a larger amount of alcohol. Shedlovsky, Ross,and Jakob (48) also have shown that a minimum surface tension of 22.6 dynes/cm -ould beobtained with sodium dodecyl sulphonate by adding 0.5 percent dodecanol.

The presence of a good solubilizing agent such as sodium lauryl sulphate greatly in-creases the concentration of a slightly soluble alcohol, such as 1-dodecanol, in an aqueoussolution. Inasmuch as a dynamic equilibrium exists between the amount of 1-dodecanoladsorbed at the interface and the amount present in solution, the use of the solubilizingagent should lead to more molecules being adsorbed, a closer packing of the hydrocarbon


Tabie 5Critical Surface Tension ( ) and Maixinmun Work of Adhesion (

of a Series of Related Low-Energy Surfaces

Atom

Atomic Percent Maximull,Surface Constitution Re'placnement (dynes,cm) (ergs cm-)

of Cl by F

Perchloropentadiicno ic 1CC 0 43 103

acid (monolayer)

Polychlorot iifluuru- -CF 2 -CFCl- 75 3 73ethylene

Polyethylene -CH 2 -CH 2 - 0 31 82

Poly(vinyl fluoride) -CH 2 -CH F- 25 28 92

Poly(vinylidene fluoride) -CH 2 -CF 2 - 50 25 88.5

Polytrifluoroethyl.ene -CHF-CF 2 - 75 22 72.5

Polytetrafluoroettylene -CF 2 -CF2- 100 18.5 58

Polyhexafluoropropylene -CF 2 -CF(CF 3 )- 50 16,2 52,5

Polymethac rylic ester -.CF 3 100 10.6 45of 1H,1H-pentadeca-fluoro- 1 -octanol

11-(pe•rfluorodecyl)unde- -CF 3 100 7.8 41canoic acid (monolayer)

chains, and thus to a lower surface tension. A surface tension of 22 to 24 dynes/cm ap-pears to be near the minimum surface-tension value that can be obtained when the ad-sorbed compound is somewhat soluble, i.e., is in equilibrium with molecules present inbulk solvent. With completely insoluble monomolecular films that can be studied on aLangmuir-Acam film balance, where one can apply external pressure to compress theadsorbed molecules into the closest possible area per molecule, it may be possible toreduce the surface tension of water somewhat further. However, the molecules in thisconfiguration would not be in an equilibrium condition.

In order to cause an even greater decrease in the surface tension of water, it isnecessary to employ a soluble surface-active agent whose hydrophobic group decreasesthe surface energy of water below 24 dynes/cm, i.e., below the value of ,. of close-packed methyl groups. It is apparent that a much greater lowering of the surface tensionshould be possible if the paraffin chains are replaced by perfluoroalkane chains, for ,of a close-packed layer of perfluoromethyl groups is much less than 24 dynes/cm. Thesame conclusion also follows from the lower intermolecular cohesive energy betweenclose-packed, adlineated perfluoroalkane chains than between adjacent alkane chains.Early hints to this effect were found on comparing the surface tensions of various or-ganic compounds with those of the analogous fully fluorinated derivatives (Table 6).

During tho early postwar years many fluorinated aliphatic compounds of highermolecular weight became available for research, and some were later put on the market

NAVAL RESEARCH LABOPATORY 15

as surface-active agents In Fig. 13 areplotted the most reliable surface -tensiondata at 20 C for aqueous iolutionF of a -,----PURE WATER

variety of compounds as a functiozi of the 70 -

concentration. Nut only is it evident thatfluorinated surface-active a .;vnts c anlower the surface tension of water at 20 60C considerably below 24 dynes cm, butthe steep initial slopes reveal h1.,v eiti-

U,

cient nianv ui theni are in accomplishingtheir tasks. Figure 13 also reveals thatnone of the agents prepared to date areable tolower the surfaeŽ tension of waterbelow 14.5 dynes.;cm at 20 C. However, 4 0

it might be possible to lower the surface /Q4. /,//'/X. /

tension even more by using appropriate /,/"• ; //synergistic combinations designed to ob- 30 - /"/•<.' :/7' /tain closer packing of tica adsorbed per-fluoroalkane chains. Where Lhe terminal,or , carbon atom of such compounds has 20

the composition of -CF 2 Hinsteadof -CF 3 ,the minimum attainable surface tens:.nis considerably higher, as is clearly II ____

shown by Figs. 14 and 15, which are taken 0 0. 0.4 0 6 08 [ 0

from a previous study (49). This result CONCENTRATION M)

agrees with what would be expected froma consideration of the relative values of Fig. 12 - Surface-tcnbiion va'ues for

for -CF 3 and -CF 2H groups (Table 2). the majority of the conventional sur-face-active agents, within shaded

burface tension versus log concen- area (43)

tration curves for many fluorochemicalsin water show discontinuities in slope thatare characteristic signs of micelle formation, and the solution concentration associatedwith the discontinuity is called the critical micelle concentration (CMC). Arrington and

Table 6Effect of Perfluorination on Surface Tension of Several Organic Liquids at 20WC

Co[pound i I Perfluorinated I ICompound (dynes/cm) Derivative (dynes/cm)

n-pentane 12.0 Iperfluoro-n-pentane 9.9

n-heptane 20.3 perfluoro-n-heptane 13.2

n-octane 21.8 pe--fluoro-n-octane 13.6

methyl cyclohexane 23.8 perfluoro - vethyl 14.7cyclohexane

n-dibutyl ether 22.9 perfluoro-n-dibutyl 13.4ether

n-butanoic acid 26.8 perfluoro-n-butanoic 16.3acid


40

4C30-

owl 0.01 1.0

Solute Concentrotlon (Weight Pereont)

Fig. 13 .. Surface tenpion vs solute concentrationcurves for a variety of perfluoro derivations inaqueous solution (numbers in parentheses indi-cate references)a. Perfluorododecanoic acid (49)b. Perfluorodecanoic acid (49,54,57)c. Perfluorooctanoic acid (49)d. Perfluorobutyric acid (49,54)e. Perfluorodecanoic acid (86)f. Perfluorooctanoic acid (54,57,58,86)

g. Perfluorohexanoic..ac id (54,57,86)h. C 7 Fs5 CONHC 2 H 4 NI) CI (F-6) (87)i. [Cs F 7 SO 2 NHC 3 H6 (CH 3 )3 1] 1- (58)j. CGF, 7 SO 2N(C 2 H,) (C 2H 4 0),4 H (58)k. C,Fz7SO2 N(C 2 Hs)C 2H4OPO(OH) 2 (58)1. CF 1 7 SOLi (58)

m. CBFIIS0 3K (58)n. C 5 F 1 7 So 3H (58)0. C7F CONHC 31iN + (CH) 2CH 4C00- (58)p. [C 7FsCONHC3 H6N(CH 3)J1] (58)r. C 7 F1 5 CONH 4 (58)s. Perfluorooctanoic Acid in 0.05M KC| (54)1. Perfluorooctanoic Acid in 0.025M IICI (54)u. P1.rfluorooctanoic Acid in 0.05M HCI (54)x. Potassium Perfluorooctanoate (54)y. C,F 17 S0 3K in 9.4M HNO 3 (88)

Patterson (50) determined the CMC of several perfluoroacids and their ammonium salts.They found that the CMC of the acids was dependent upon chain length and was lower thanthe CMC of the corresponding hydrocarbon acid. Finally, they reported that the silts ofthe acids were more soluble than the free acids and had higher CMC values.

Klevens and co-workers (51-55) confirmed the presence In aqueous solution ofmicelles of perfluoroacids, as well as other fluorochemical derivatives, by utilizing avariety of surface-chemical techniques including electrical conductivity, pH, surfacepotential, rurface tension, and dye-adsorption measurements. Their results are sum-marized in Table 7. Values of the CMC obtained by the dye-adsorption technique are allFomewhat smaller than those determined from surface-tension measurements, but are


1111111 , � 111111111

O.PCRVLuOROSUTYRIC ACIDD.PCRVLuoNOOCTANoIC ACID

* AsPERPLUONOOECANOIC ACID60 �*PIftFLUON0OODICANOIC ACID

* * * SCHOLSEAG �1)U*KLEVENS (54)

S� 50

ci

wZ4a

�3O 0I-maU4

� 20S

U

IC iitil I I II iiiil liii iiil I I III iiil I 111110001 0.01 0.1 1.0 10.0 100.0

SOLUTE CONCENTRATION (WEIGHT PERCENT)

Fig. 14 - Surface tensions of perfluoro acidsin aqueous solutions

7C I I ii,,,

O * H(CF2 )COONN 4o * H(CF3 )*COOII

A' H(CF2),COONN 4z; 60 �'H(CF 2 )gC00N6' PATTERSON (50)

I-4� 50U

0U2

O 402 52('Izmaa- 30ma

4C.) Sa'-

L'� 20

0.01 0.1 1.0 10.0 1000

SOIUTE CONCENTRATION (WEIGHT PERCENT)

Fig. 15 - Surface tensions of �-monohydro-perfluoro compounds in-aqueous solutions

--

18 NAVAl RESEARCH tABORAlORY

Table 7A Comparison of the Critical Micelle Concentrations (CMC) of Fluoracids and Po-

tassium Fatty Acid Soaps as Determined by the Spectral-Dye Method*(T = 25-C)

Critical Micelle Concentrations (moles per liter)

Number -

C atoms F -Hydro-in chain Fatty acid Perfluoro- Kel-F perfluoro-

soaps acids acids' acids

2 2.06

3 1.11

4 0.53 0.70

5

6 1.68 0.054 0.062

7 0.75 0.15

8 0.39 0.0056 0.0091

9 0.20 0.03

10 0.095 0.00048

11 0.050

12 0.025

13 0.012

14 0,0058

*From Ref. 55.tCl(CFCFCI)xCFCOOH (X 1, 2, and 3)

roughly equivalent to the CMC values determined by conductivity studies. The data inTable 7 show that perfluoroacids have much lower CMC values than either the corre-sponding unfluorinated acids or the . -hydroperfluoroacids. It is also worthy of note thatthere is micelle formation even in perfluoroacetic acid, whereas a four-carbon chain atleast is required for micelle formation in salts of the fatty acid (53),

Offsetting their greater surface-tension-lowering ability than convent4.onal surface-active agents, the perfluoroaliphatic compounds have the disadvantage of inuch higherunit cost. A way to decrease this limitation was revealed when Bernett and Zisman re-ported (56) the existence of synergistic combinations of a, conventional water-solubleagent and a perfluoroalkyl surface-active agent. In these combinations the conventionalsurface-active agent is believed to serve as the major water-soluble micelle-forming(or solubilizing) ingredient, and the much less soluble perfluoroalkyl compound is theminor ingredient which is solubilized in the hetero (or mixed) micelles. The dynamicexchange of the perfluoroalkyl compound between the heterc micelles and solution keepsthe water-air interface covered with an adsorbed monolayer, one essentially comprised


of the perfluoroalkyl compounds: hence, the very low surface tensions characteristic ofperfluoro surface-active agents can be obtained with much smaller concentration than ifa more soluble homologue were used.

Many applications of these new agents have been reported (57,58), and more new usesare being found rapidly. Low-energy solid surfaces such as the polypropylenes, fluorine-containing plastics such as poly(vinyl fluoride), and various solids coated with a mono-layer of a fatty acid (or derivative) or with a silicone have become increasingly morecommon. Whenever such materials must be completely wetted by water (or water emul-sions), difficulty is often encountered with solutions of conventional wetting agents. Awetting agent derived from silicones or from fluorinated aliphatic compounds will usuallybe more effective (49,56). Since many perfluoroalkyl derivatives are able to lower thesurface tension of water below the ), value of any hydrocarbon surface, such solutionscan spread spontaneously on oil or gasoline. This fact is basic to the recent NRL develop-ment by Tuve, et al. (59) of a new and impressive water foam fire-fighting technique forfuel fires during aircraft disasters.

INSOLUBLE FLUOROCHEMICAL MONOLAYERS AT THE LIQUID/GAS INTERFACE

In their early investigation of monolayers of fluoroacids and alcohols spread onaqueous substrates, Arrington and Patterson (50) reported that H(CF 2 )1 0 CH 2OH could bespread on water to give an insoluble monolayer. From the curve of film pressure (F) vsarea per molecule (A), obtained using a conventional Langmuir-Adam hydrophil balance,they determined that the area per adsorbed molecule at closest packing was 25 A 2 . (Inclose agreement is the value of 25.1 k2 obtained by Bernett and Zisman (25) from mea-surement on a Stuart-Briegleb ball model.) Fluoroalcohols of lower molecular weightcould not be spread successfully by Arrington and Patterson, presumably because oftheir solubility. Monolayers of the fluorocarboxylic acid H(CF 2)1 2 COOH on 0.01N HC1behaved normally, but on water gave evidence of being unstable. Because their fluoro-cat-boxylic acid monolayer on 0.01N HCI behaved as a perfect two-dimensional gas up tofilm pressui-es as high as 5 or 6 dynes/cm, they concluded that intermolecular forces ofattraction between adjacent fluorocarbon chains were weaker than those between the hy-drocarbon chains.

H. W. Fox (60) studied at NRL the behavior of insoluble monolayers of .- trifluoro-stearic acid and,. -trifluorooctadecylamine on aqueous substrates of various pH and foundthese monolayers were more expanded and less stable than monolayers of the unfluorin-ated compounds. He attributed the instability to the mutual repulsion of the strong dipolesassociated with the ,,-trifluoro groups. The large dipolu moment of the -CF 3 group alsogreatly reduced the surface potential of the adsorbed monolayer. The maximum surfacepotential difference of ,.-trifluorostearic acid on pH 2.5 was -705 millivolts, which is tobe compared with a value of +355 millivoltn for stearic acid.

Klevens and Davies (52), from a study of F vs A and surface potential difference(.V) vs A for monolayers of C9 F1 9 COOH on aqueous substrates, determined that theenergy of adsorption of a -CF 2- group at a water/air interface was at least twice that ofa -CH 2- group (1490 cal/mole compared with 725 cal/mole). They also reported that theclosely packed acid film caused a remarkably low surface-potential change of AV = -1000millivolts; this value was reported to be independent of substrate pH from pH 2 to pH 12.

Bernett and Zisman (25) determined F vs A and AV vs A curves for two homologousserics of progressively fluorinated fatty acids spread as insoluble monolayers at thewater/air interface. These compounds were the same preparations made by Braze (21)which were described previously in this report as the "segmented" acids. Both seriesof acids exhibited monolayer stability which increased with increasing length of the fluo-rocarbon chain. A chain length of at least 18 carbon atoms was necessary in both series


of acids for the formation of stable monomolecular films. None of the progressivelyfluorinated acids produced solid, condensed monolayers. Presumably, that was a resultof steric hindrance to chain alignment, since the fluoroalkyl chain has a cross-sectionalarea roughly 35 percent greater than the unfluorin~ted alkyl chain to which it is attached.The closest molecular packing observed was 27.5 A2 per molecule, which is 10 percentgreater than the cross-sectional area of the perfluoroalkane chain. Changes in surfacepotential caused by these adsorbed fluorocarboxylic acid monolayers were also extremelylow; for example, ll-(perfluorodecyl)-undecanoic acid caused a 'V of -815 millivolts atclosest packing on 0.01 N H 2SO7.

Bernett, Jarvis, and Zisman (61) have reported that at high film pressures the F-vs-A curves of 18-fluoro-l-octadecanol were identical with those of 1-octadecanol, but thatat low film pressures the 18-fluoro alcohol was somewhat more expanded. The film ex-pansion was again attributed to the large dipole moment of the C-F bond, which leads atclose packing to strong repulsion between the similarly oriented dipoles in the terminal-CH 2 F groups. Molecular ball model calculations showed that a single fluorine atom inthe omega position was too small to sterically hinder packing of the hydrocarbon chain.Only with the larger halogens, iodine and bromine, did the steric factor become signifi-cant. The large C-F dipole moment also caused 18-fluoro-l-octadecanol to have a muchlower value of AV than octadecanol (-55 millivolts as compared with 620 millivolts).

Ellison and Zisman (62,63) found that it was possible to spread certain fluorocarbonderivatives, as well as silicones and organosilicates, as insoluble monomolecular filmson a variety of organic liquids. Using an all-PTFE (polytetrafluoroethylene) version ofLangmuir's early film balance, they spread nionolayers on such liquids as hexadecane,white mineral oil, and tricresyl phosphate. They proved that these films were oriented,adsorbed monomolecular films which obeyed the same surface-chemistry laws alreadyestablished for conventional surface-active agents adsorbed at the water-air interface.Their work clearly pointed to the promise of high fluorochemical surface activity withrespect to a variety of organic liquids, but the need was also obvious for a study of thesurface tensions of solutions of similar compounds.

SURFACE ACTIVITY OF FLUOROCHEMICALS IN NONAQUEOUS SOLUTIONS

In seeking compounds which would be surface active when dissolved In nonaqueousliquids, it was realized that the mechanism causing solute adsorption should be most ef-fective when the adsorbed film caused the surface-energy decrease to be a maximum.Therefore, for a given solute there should be a greater surface-active effect, the higherthe solvent surface tension. For this reason, a great variety of common polar-nonpolarorganic compounds are surface-active agents in water because water has the highestsurface tension of all common liquids at ordinary temperatures (72.8 dynes/cm at 20'C).Most of the liquid organic compounds, however, have surface tensions at 20'C between30 and 45 dynes/cm. Thus it should be difficult to find an effective surface-active solutefor hexadecane, because its surface tension is only 26.7 dynes/cm at 20°C. One mustturn to derivatives of the silicones or highly fluorinated hydrocarbons for suitable solutesfor an organic liquid having such a low surface tension. In turn, one would be hardpressed to find a suitable surface-active agent for use in perfluorocarbon solvent.

How should a surface-active molecule be designed so that it will be an efficient surface-tension depressant for organic liquids? An answer has been found (64-68) by reason-ing as follows. It is well known that surface-active agents for water systems have acharacteristic hydrophobic-hydrophilic structure such that the hydrophilic portion of themolecule can be dissolved in or nsociated with the water in the bulk or surface phase,with the hydrophobic portion of the molecule remaining essentially out of the water.Similarly, a surface-active agent for an organic liquid should have an asymmetric struc-ture such that one extreme portion of the molecule is made up of one or more organophobic


griuups and the opposite extreme portion is made up CH 1COOCH 2(CFP)*CF,of one or more organophilic groups. Structural dia- CHCOOCHZ(CF,)*CF3 Igrams of three of the many such organic structures Isynthesized at NRL by O'Rear and Sniegoski (69) are CHCO0CN2 (CF,), CF,given in Fig. 16. For example, in the third compoundthere are two organophobic CFj(CF 2 ),CH2- groups. cH2 CHZ OOC(CF2 )2 CF3Theorganophilic portionof the moleculeopposed to it Iis a benzene ring containing two ester groups andfour covalent chlorine atoms. Chlorine was also de- CH2 CNH OOCCFW)2 CF3sirable in this structure to decrease Its volatility.Incidentally, this molecular design Is an application Clof the previously mentioned conclusion that covalent Cl . COOCH2 (CF2 )s CF3chlorine substit'ition has an effect on the surfaceenergy of the molecule opposite to covalent fluorine. CI COOCH 2(CF2 )s CF3

Some of the interesting results of the investiga- C1

tion of these compounds in various organic liquids Fig. It, - Structure o fare summarized in Figs. 17, 18, 19, and 20. Thus, several fluorine-con-Fig. 17 illustrates the fact that the higher the sur- taining compounds usedface tension of the solvent, the greater the surface as additives for certainactivity of a given solute. The higher surface ac- organic liquids and poly-tivity is indicated by the steeper initial slopes. Fig- mersure 18 illustrates the effect of varying the surface-active solute on the surface tension vs concentrationcurves in propylene carbonate. Figure 19 is the result of using the Gibbs' adsorption re-lation to compute from the surface-tension-vs-concentration curves in Fig. 18 the ad-sorption of the surface-active agents. Figure 20 is a plot obtained from the data in Fig.18 in terms of F, the film pressure (total surface tension lowering), vs the area A occu-pied per adsorbed nolecule. These calculations are based on the assumption that thesolute adsorbs as a monomolecular film. These F-vs-A graphs are very much like thoseobcained with conventional surface-active agents adsorbed at the water-air interface.Finally, in Fig. 21 are the results of fitting the previous curves to Langmuir's early idealgas-law relation (70) in two dimensions

*30

NY AROCO4 1240442 )

0,2

20

IN | / •,ACM •. ROUTADIEN

1 101

r I / IN N-HEXADECANE

0 I 2 3 4 5 6 7 aSOLTiE CONCENTRATION (WT PERCENT)-•

Fig. 17 - Adsorption exccsb vu concen-trations of 111,1 11,9H-hexadecafiuoro-nonyl butyrate in various solvents (alldata at 205 C)


48 42

46- 40

44 PROPYLENE CARBONATE 38

42 36

40 34

-38 32

u 36 30S34 -e

S32 -26

30 0-24

28 V-

26 --

24

22-2ol I I I I I 1 1 _L _

0 001 002 003 004 005 006 007 008 009 010CONCENTRATION (MOLES/LITER)

Fig. 18 - Surface tensions of partially fluori-nated compounds in propylene carbonate: ',1 ,H, lIl-pentadecafluorooctyl alcohol; A. IH, IH-pentadecafluorooctyl ethanesulfonate; v, bis-(Ili, IH - pentadecafluorooctoxy)-bis -(t-butoxy)silane; Y, tris (1KH,!1 H-pentadecafluorooctoxy)-t-butoxysilane ;I , bis-( 111,1K- pentadecafluor-ooctyl) tetrachlorophthalate; ,, tris - (IH, IIIheptafluorobutyl) tricarballylate; 0, tris-( li-,I H-pentadecafluorooctyl) tricarballylate; *,bis - (!11,111 - pentadecafluorooctyl) dodecenylsuccinate; c, bib -( 1i, 1I l-pentadecafluorooctyl)toluene dicarbainate; x, hexyl perfluorobuty-rate; C., 1,2.3-trimethylolpropane tris (per-fluorobutyrate); *, .bis -(ll, 11-1-undecafluoro-hexyl)-3-methylglutaratv; w, bis-(lll, lf-pen-tadecafluorooctyl) 3-methylglutarate

I V F f ) I- .1 , A. T (5)

and to the relation obtained recently by Hedge (71) from the more general equation of state

derived for gaseous monolayers on water by Phillips and Rideal (72).

I - .1, kT. (6)

If one considers a plastic solid to be either a supercooled liquid or a liquid of ex-tremely high viscosity, one would expect many partially fluorinated compounds to showsurface activity when dissolved in various polymers. The rate at which these surface-active agents adsorbwill be primarily determined by the rate of diffusion in the bulk polymerphase. Jarvis, Fox, and Zisman (73) found that a variety of fluorine-containing additives,such as those listed in Fig. 16, would adsorb very strongly at the polymer-air interfacewhen incorporated in the bulk polymer. It was shown that the additives upon adsorptioncaused remarkable changes in the surface properties of the polymer. Just 0.5 percent oftris(l i,1lK-pentadecafluorooctyl) tricarballylate reduced the ., of poly(methyl methacry-late) from 40 dynes/cm to less than 19 dynes/cm, approximately that of a polytetrafluoro-ethylene surface.

In addition to reducing the oil and water wettability of polymer surfaces, the adsorbediluorochemical additives should also influence the surface in other ways, such as reducingadhesion and reducing its coefficient of friction. In this regard, [kowers, Jarvis, and


PHOPYLENE CARBONATE 26

26

e4 24

22 -

- 0 20

18 187. .6, 16

14 0014

S12 12

101 10

88

66

4 4

2 2

000 0' 002 003 0 00005 0004

CONCENTRATION (MOLES/LITER)

Fig. 19 l Surface excess of partially fluorinatedcompounds in propylene carbonate: A, IHIH-pentadecafluorooctyl alcohol; A , 1HIH-penta-decafluorooctyl ethanesulfonate; V, bis-( I- IlH-pentadecafluorooctoxy)-bis -(t-butoxy) s iI an e;V-!, bis (Ill, lH-pentadecafhtorooctyl) tetra-

chlorophthalate; o. tris-(IH, IH-heptafluoro-butyl) tricarballylate; 0, tris-(iH, IH-penta-decafluorooctyl) tricarballylate; l, bis-(lI1l, IH-pentadecafluorooctyl) dodecenyl succinate; I,bis-(IH, 111-pentadecafltorooctyl) toluene di-,.arbamatc

Zisman (74) reported that the coefficient of friction of poly(methyl methacrylate) andseveral of its copolymers with vinylidene chloride could be reduced to 0.10 or less by theincorporation of only 1 percent of a suitable fluorochemical additive in molded polymerdiscs. An interesting and important property of these systems is the ability of the adsorbedfluorocarbon film to be self-repairing. If the initially adsorbed fluorocarbon film isremoved, other molecules of the surface-active additive will diffuse into the interface andrestore the original surface behavior.

ADHESIVE AND ABHESIVI- BEHAVIOR OF FLUOROCHEMICALS

In ancient times it was found that solids could be made to adhere strongly by wettingeach surface to be joined with a thin layer of a liquid which hardened or solidified gradu-ally during contact. Until this century the selection of suitable adhesives and applicationtechniques was an art, depending on the use of glue formulations made from fish andanimal products or cements made from Inorganic slurries or solutions.

Pioneering studies of adhesion by McBain and Hopkins (75) led them to report that"Any fluid which wets a particular surface and which Is then converted Into a tenaciousmass by cooling, evaporation, oxidation, etc., must be regarded as an adhesive for thatsurface." McBaln and Lee (76) soon afterward added the requirement that the adhesivemust be able to deform during its solidification to release elastic stresses developed inthe forming of the joint. Their work made evident the following three requirements foran adhesive: (a) wetting, (b) solidification, and (c) sufficient deformability to reduce thebuildup of elastic stresses in the formation of the joint. Hull and Burger (77) found the

24 NAVAL RESFARCH LABORATORY

15

14

03

IQ PROPYLENE CARBONATE

I'

IC

S7g6

3

2

50 70 90 ItO 130 150 170 190 210 23o 250AREA/lMOLECULE (12)

Fig. 20 - Force-area isotherms of partiallyfluorinated compounds in propylene carbonate:"1, 11, IH-pcntadecafluorooctyl alcohol; A,

IF!, 1tl-pentadecafluorooctyl ethane sulfonate;., bis -(111, 1 1-pentaclecafluorooctoxy)-bi s -(t-butoxy)-silane;! ], bis( If. I I -pntadecafluoro-octyl) tetrachlorophthalatt,; , tris-(llt, ll-heptafhlLorobutyl) tricarballylate; 6, tris-(Ill,IIl-pentadecafluorooctyl) tricarballylate; E,bis-(!It,1 lI-pentadteca fluorooctyl) dodec.nylsuccinate; C. bis-( 111, ll.-pentaducafhitoro-octyl) toluene dicarbanate

same rules necessary for good glass-to-metal joints. The second and third require-ments have been fully investigated and reported. The first requirement was studiedmuch later by deBruyne (78) and Zlsman (79,80).

When a liquid adhesive solidifies, the reversible work of adhesion of the adhesive tothe adherend would still be close to the value computed in the calculations shown in Eqs.(5) and (6) for the adhesive in the liquid state if it were not for the development of stressconcentrations. This conclusion follows from the highly localized nature of the attractivefield of force causing adhesion. Since this attractive force is effective little more thanthe depth of one molecule in both the adhesive and the adherend, it will be unaffected bychanges of state so long as allowance Is made for any resulting alteration in the surfacedensity or molecular orientation occurring at the joint interface (79,80). The former canbe estimated from the change of density on solidification, but the latter may be difficultto compute, since reorientation effects could originate through a crystallization processstarting from some nucleus not located in the interface. Internal stresses and stressconcentrations usually develop on solidification of the adhesive, the most common causebeing the difference in the thermal expansion coefficients of the adhesive and adherend.In many applications of adhesives this matching process is not readily done or is neglectedbecause it is not critical. However, In general, the theoretical strength of the adhesivejoint is considerably decreased by the development of internal stress concentrations.

Effects of internal stresses developed on solidification and of stress concentrationsduring applications of a joint have been given much attention; an excellent discussion ofthis area of research has been given by Sneddon (81). Photoelastic studies by Mylonas(82) and Mylonas and deBruyne (83) led to the conclusion that in a lap joint, poor wetting


9 -0- EXPERIMENTAL DATA4 -6-(F4 0 /A2 )(A-A.)-KT

S7 -0-(F+F.)(A-A.)-KTU,uZJz 6-

0 5-

3

2

I0 ------ -I I I I l I I I I40 60 80 100 120 140 160 180

ABEA/MOLECULE (CA

Fig. 21 - Force-area isotherms of bis-(1i1, 1 H - heptafluorobutyl)-3-methylglu-tarate adsorbed at the propylene carbon-ate-air interface

of the adherend tends to produce a greater stress concentration at the free surface of theadhesve where failure is most likely to be initiated. As the contact angles become large,the maximum stress concentration increases and moves toward the lineal boundary,where the adhesive and atmosphere make contact with the adherend, the stress concentra-tion factor increasing from about 1.2 to 2.5. Griffith (84) had shown that failure of theadhesive may occur at a relatively small applied stress if there are bubbles, voids, in-clusions, or surface defects; it occurs because stress concentrations result which aremuch higher than the mean stress applied across the specimen.

Griffith's conclusion is especially important when considered in terms of the probableeffect of poor wetting on the development of air pockets at the adhesive-adherend inter-face. It has already been shown that if = 0, the theoretical joint strength far exceeds thetensile strength of most adhesives. In practice, the theoretical joint strength is not at-tained; evidently, one major cause is the development of stress concentrations duringsolidification of the adhesive. When . = 0, there may be gas pockets formed at the adhe-sive-adherend interface around which stress concentrations can build up, for if the adhe-sive is too viscous when applied, It may never penetrate the accessible surface poresbefore polymerizing. Of course, this situation is aggravated if ý - 0.

In adhesives technology it is common practice to roughen the surface of each ad-herend, or "give it tooth," and so obtain a stronger joint. This practice can be justifiedtheoretically with reservations by the following considerations. If gas pockets or voidsin the surface depressions of the adherend are all nearly in the same plane and are notfar apart (as on the upper adherencf A Fig. 22), there may be crack propagation from onepocket to the next, and the joint may break as if it has a built-in "zipper." Therefore, ifroughness must be accepted, the kind of roughness shown nn the lower adherend would bepreferable because crack propagation along a plane would be less probable (79,80).

An adequate theory of adhesive action should be able to explain poor as well as goodadhesion; hence, a test of any theory is. its ability to explain the properties of abhesives(79,80). "Abhesives" are films or coatings which are applied to one solid to prevent-orgreatly decrease-the adhesion to another solid when brought into intimate contact with it.Abheslves are employed in molding, casting, or rolling operations; therefore, it is common


F' SMOOTH ADHEREND

GAS BUBBLES ADHESIVE

/ ROUGH ADHEREND/ / / '/ I

Fig. ,, - Effect of surface roughness oncoplanarity of gas bubbles

to refer to the film as the parting, mold-release, or antistick agent. Examples of ma-terials commonly used for such purposes are the polymethylsiloxanes, the high-molecular-weight fatty acids, amines, amides, and alcohols; various types of highly fluorinated fattyacids and alcohols; and polytetrafluoroethylene or films or copolymers of polytetra-fluoroethylene and hexafluoropropylene deposited by coating the mold with an aqueous dis-persion of the polymer, drying it, and finishing with a brief bake at a high temperature.Usually a condensed monolayer of the agent is sufficient to cause the optimum partingeffect. Table 4 gives the values of , of each of these films. Evidently, each abhesivecoating converts the solid into a low-energy surfaze, and any liquid placed on such acoated surface will exhibit an equilibrium contact angle which will be larger asincreases. When r, is large enough, such poor adhesion results that the application ofa modest external stress suffices for effective mold release. However, the excellent andeasy parting action observed needs additional explanation.

It was pointed out earlier that the value of it, for the various low-energy surfacesreported here exhibits a maximum variation of threefold in going from the least to themost adhesive materials. This range is much too small to cxplain the effectiveness andeasy parting action caused by a film of a good abhesive. Note that the value of it' is fora flat, nonporous, smooth surface. The roughness factor r of the uncoated surface of themold could suffice to raise ItI by a factor of from 1.5 to 3 or more, depending on thesurface finish; however, if the material to be molded Is (or becomes) viscous rapidlyduring injection or application, poor wetting will cause pockets to be produced at the in-terface between the material molded and the abhesive, and thereby the adhesion will begreatly decreased through stress concentrations by some unknown fraction l/g. Also,when - -30', the resulting stress-concentration factor of from 1.2 to 2.5 at the abhesive-plastic interface acts to decrease adhesion a, "ordingly. Therefore, the adhesional workper unit area of the apparent or envelope area at the molding interface will be fr ,, i LrFor the best abhesive action, the release agent should have the larg'-st possible value ofI ror the substances to be molded against it. The large values of encountered with thevarious types of organic liquids on such low-energy surfaces as those shown in Table 4make it evident why these film-forming materials are so effective as abhesives. As ageneral proposition, any low-energy surface will be more effective as a release agent (orabhesive), the lowei its value of . or the lower it 1 (79,80). It also can be concludedfrom the preceding discussion that the smoother the finish of the outer coating of the ab-hesive, and the lower the surface tension and viscosity of the material being molded (orthe lower its contact angle with the adhesive), the greater will be the external stress re-quired to cause the parting action.

Evidently, for optimum joint strength, it is essential to keep the contact angle betweenthe liquid adhesive and the adherend as small as possible in order to minimize buildup ofstress concentrations and to obtain good spreading into pores and cracks. In applying the


liquid adhesive. its viscosity should b~e as low as possil le to increase the rate ofspreading and penetration before polymerization causes solidification. Maximunm spread-ing and capillarity will be obtained with adhesives having the highest surface tensi'nscompatible with obtaining a zero (or small) contact angle. When conditions of completewettintr and freedom from the formation at the interface of gas (or vacuum) potc kets andocclusions prevail, adhesion will usually be ample, and joint failures generally will becohesive. But because pores and cracks alwý,ys exist in the surfaces of real solids, sur-face occlusions and pockets will always be formed to some extent on applying the adheive:the resulting loss in joint strength can be su large as to allow a failurc in adhesion tooccur-as happens readily in the presence of an adsorbed parting agent. It is possiblethat thv strongest joints obtainable in practice will be found when the surface tension ofthe adhesive is slightly less than the value of of the adherend. In other words, moremay be gained by minimizing the tundency to form interfacial occlusions than by maximiz-ing. ýhe specific adhesion it i.

An adhesive with a very low >, would seem to be advantageous for most adherends.Inasmuch as fluorination of a polymer lowers its liquid surface tension, Sharpe andSchcnhorn (85) recently concluded that a highly fluorine-substituted polymer would be anffl purpose adhesive. As a matter of fact, some of the copolymers of tetrafluoroethyleneand hexafluoroethylene have low melt viscosities and hence have been used for someyears as adhesives for highly fluorinated plastics like polytetrafluoroethylenes. Un-fortunately, as the surface tension of the liquid adhesive is made lower by fluorination,its cohesive strength (it ) is also lowered, since it . .. Thus the strength of thesolid is also lowered. If the adhesive is prepared as a thermosetting polymer, or be-comes highly crystLline on solidifying, some cohesive strength is thereby gained. Whenthe adherend is a highly fluorinated polymeric solid, its cohesive strength will usually notbe so great as to necessitate use of an adhesive of great cohesive strength. But if the ad-herend is a plastic or other solid of high strength, the use of a low-strength adhesive istoo limiting. Therefore, there are serious problems to be overcome in making anygeneral-purpose adhesive from high fluorinated polymers.

There has been much discussion about the necessity for using adhesives capable ofchemically combining with the adherend, but the energy of _dsorption involved in thephysical adsorption to the adhc: end of the adhesive molecules has been proved more thansufficient to form joints which are stronger than the cohesive strength of the solidifiedbulk adhesive. Chemical bonding may prove desirable in order to obtain greater heat,water, or chemical resistance, but it should not be necessary to increase the specificadhesion. However, it is usually a great advantage to be released fr.om the very limitingrequirement of finding an adhesive capable of chemically combining with the adherends.

CONCLUSIONS

This very condensed review of the surface properties of fluorocheinicals has led to anumber of generalizations which deserve a brief summation here.

1. The most nonwettable and nonadhesive solid is one whose surface consists of fullyfluorinated aliphatic hydrocarbons, the composition of the outermost layer being made upof perfluoromethyl groups in closest packing.

2. Langmuir's principle of localized action, which he applied to conventional hydro-carbon monolayers many years ago, has proved true provided that ions or unbalai.ceddipoles near the surface are avoided.

3. Whereas the polar hydrophilic group of a "conventional" surface-active agent de-termines its solubility, ionizability, and adsorptivity, it is the hydrophobic portion of themoleclue that determines the minimum surface energy it can impart to a. liquid in whichit is dissolved or to a solid surface on which it is adsorbed.

0*- --- 1w-- -- - . _. -_W_-- ~-


4. An analogous rule to the preceding one holds for the surface-active agents usedin org-anic solvents which have perfluoroalkyl organophobic groups arranged in opposi-tion to organophilic groups (such as hydrocarbon, ester, or chlorocarbon groups).

5. Perfluoroalkyl derivatives are the most surface active of all materials yetstudied, whether relative to water or other liquid systems.

6. A great future is predicted for surface-active materials made from partiallyfluorinated compounds, and also for highly fluorine-substituted monomers, and polymers,especially for protective coatings.

7. The extreme surface properties of perfluoro aliphatic compounds arise from theextraordinarily weak field of force existing in the vicinity of covalently bonded fluorineatoms and their neighbors. Although this result agrees with the fact that fluorine hasthe lowest polarizability of the halogens, the fact that the field of force is so muchweaker, even than that around the .covalent hydrogen, deserves more investigation. Prob-ably the explanation is in the values of the octapole moment or higher terms of the co-h.esive field u force rather than simply the polarizability. Another way of stating the caseis that the local attractive fields of force giving rise to cohesive effects in perfluorome-thane, the perfluoromethyl group, and in the perfluoromethylene group, are more likethose of the inert gases than the fields of force around the CH 4 ,-CH 3, and -CH 2 - groups,respectively.

REFERENCES

1. Zisman, W.A., "Relation of Equilibrium Contact Angle to Liquid and Solid Constitu-tion," Advances in Chemistry, No. 43, Amer. Chem. Soc., Washing-tn, D.C., p. 1,1964

2. Schulman, F.. and Zisman, W.A., J. Am. Chem. Soc. 74:2123 (1952)

3. Schulman, F., and Zisman, W.A., ., Colloid Sci. 7:465 (1952)

4. Ellison, A.H., Fox, H.W., and Zisman, W.A., J. Phys. Chem. 57:622 (1953)

5. Fox, H.W., Hare, E.F., and Zisman, W.A., J. Phys. Chem. 59:1097 (1955)

6. Simons, J.H., "Fluorine Chemistry," New York: Academic Press, Vol. I, p. 144,1950; Vol. II, p. 340, 1954

7. Hanford, W.E., and Joyce, R.M., U.S. Patent 2,440,800 (May 4, 1948)

8. Berry, K.L., U.S. Patent 2,559,629 (July 10, 1951)

9. Grajeck, E.J., and Peterson, W.H., American Dyestuff Reporter 6:37 (1959)

10. Grajeck, E.J., and Peterson, W.H., Textile Research J. 32:320 (19u2)

11. DeMarco, C.G., McQuade, A.J., and Kennedy, S.J., Modern Textiles 2:50 (1960)

12. Piedmont Section, A.A.T.C.C., American Dyestuff Reporter 52:25 (1963)

13. Ernlund, J.H., TAPPI 40:3 (1957)

14. Olson, M.H., Leather and Shoes 14:24 (1961)

15. Bernett, M.K., and Zisman, W.A., "Prevention of Liquid Spreading or Creeping,"Advances in Chemistry No. 43, Amer. Chem. Soc.., Washington, D.C., p. 332, 1964

16. Bernett, M.K., and Zisman, W.A., "Hydrophobic and Oleophobic FluorooolymerCoatings of Extremely Low Surface Energy," NRL Report 6039, Feb. 1964

17. FitzSimmons, V.G., and Romans, J.B., "Some Factors Influencing the Life and Per-formance Reliability of High-Precision Potentiometers," NRL Report 6287, May 12,1965

18. FltzSimmons, V.G., Romans, J.B., and Singleterry, C.R., "A New Approach to Lubri-cating Ball Bearings," NRL Report to be published

19. Shafrin, E.G., and Zisman, W.A., J. Phys. Chem. 64:519 (1960)

20. Gavlin, G., and Maguire, R.G., J. Org. Chem. 21:1342 (1956)

21. Brace, N.O., J. Org. Chem. 27:4491 (1962)

29


22. Shafrin, E.G.. and Zisman, W.A., J. Phys. Chem. 66:740 (1962)

23. Gent, W.L.G.. Quart. Rev. (London) 2:383 (1948)

24. Shulman, H.G., Dailey, B.P., and Townes, C.H., Phys. Rev. 78:145 (1950)

25. Bernett, M.K., and Zisman, W.A., J. Phys. Chem. 67:1534 (1963)

26. Hardy, W.B., "Coilected Scientific Papers,' Cambridge University Press, Cambridge,1936

27. Levine, 0., and Zisman, W.A., J. Phys. Chem. 61:1068 (1957)

28. Levine, 0., and Zisman, W.A., J. Phys. Chem. 61:1188 (1957)

29. Bewig, K.W., and Zisman, W.A., "Low Energy Reference Electrodes for Investigat-ing Adsorption by Contact Potential Measurements," Advances in Chemistry, No. 33,Amer. Chem. Soc., Washington, D.C., p. 100, 1961

30. Shafrin, E.G., and Zisman, W.A., "Hydrophobic Monolayers and Their Adsorptionfrom Aqueous Solution," in "Monomolecular Layers," edited by H. Sobotka, Amer.Assoc. Adv. Sci., Washington, D.C., 1954

31. Fox, H.W., Hare, E.F., and Zisman, W.A., J. Colloid Sci. 8:194 (1953)

32. Hare, E.F., Shafrin, E.G., and Zisnian, W.A., J. Phys. Chem. 58:236 (1954)

33. Cottington, R.L., Shafrin, E.G., and Zisman, W.A., J. Phys. Chem. 62:513 (1958)

34. Ellison, A.H., and Zisman, W.A., J. Phys. Chem. 58:260 (1954)

35. Shafrin, E.G., and Zisman, W A., J. Phys. Chem. 61:1046 (1957)

36. Shepard, J.W., and Ryan, J.P., J. Phys. Chem. 63:1729 (1959)

37. Bowden, F.P., anid Tabor, D., "The Friction and Lubrication of Solids. Part II,"London: Oxford University Press, 1964

38. Bangham, D.H., Trans. Faraday Soc. 33:305 (1937)

39. Bangham, D.H., and Razouk, R.I., Trans. Faraday Soc. 33:1459 (1937)

40. Martinet, J.M., Rapport C.E.A. 888, Centres d'Etudes Nucleaires deSocloy, 1958

41. Graham, D., J. Phys. Chem. 66:1815 (1962)

42. Gibbs, J.W., Trans. Connecticut Acad. 3(1876-1878), "Collected Works," Vol. 1, NewYork: Longmans, Green and Co., 1928

43. Fischer, E K., and Gans, D.M., Ann. N.Y. Acad. Sci. 46:371 (1946)

44. Jarvis, N.L., and Zisman, W.A., "The Stability and Suriace Tensibn of Teflon Disper-

sions in Water," NRL Report 5306, May 1959

45. Miles, G.D., and Shedlovsky, L., J. Phys. Chem. 48:57 (1944)

46. Burcik, E.J., and Newman, R.C., J. Colloid Sci. 9:498 (1954)


47. Harva, 0., Rec. trav. chim. 75:101 (1956)

48. Shedlovsky, L., Ross, J., and Jakob. C.W.. J. Colloid Sci. 4:25 (1949)

49. Bernett, M.K.. and Zismin, W.A.. ,. Phys. Chen. 63:1911 (1959)

50. Arrington, C.H., Jr.. and Patterson, G.D.. J. Phys. Chem. 57:247 (1953)

51. Klivens, H.B., and Vergnulle. I., lind Internationial Cungrt'ss of Surface Activity.London. England, April 1957, Vol. 1, Gas/Liquid and Liquid, Liquid Interlace.Academic Press, New York, pp. 395-404. 1957

52. Klevenis, H.B., and Davies. J.T., IlInd International Congress of Surface Activity,London, England, April 1957, Vol. I. Gas, Liquid and Liquid/ Liquid Interface,Academic Press, New York, pp. 31-35, 1957

53. Kleveris, H.B., and Raison, M., Reports of the 1st World Cona.ress on Surface Active

Agents, Paris, Section 1, p. 50 (1954)

54. Klevens, H.B., and Raison, M., J. Chim. Phys. 51:1 (1954)

55. Klevens, H.B., Koll. Zeit. 158:53 (1958)

56. Bernett, M.K., and Zisman, W.A., J. Phys. Cheni. 65:448 (1961)

57. Scholberg, H.M., Guenthner, R.A., and Coon, R.I., J. Phys. Chem. 57:923 (1953)

58. Guenthnet, R.A., and Vietor, M.L., 1. & EC Product Res. and Develop. 1:165 (Sept.1962)

59. Tuve, R.L_, Peterson, H.B., Jablonski, E.J., and Neill, R.R., "A New Vapor-SecuringAgent for Flammable-Liquid Fire Extinguishment," NFL Report 6057, Mar. 13, 1964

60. Fox, H.W., J. Phys. Chem. 61:1058 (1957)

61. Bernett, M.K., Jarvis, N.L., and Zisman, W.A., J. Phys. Chem. 68:3520 (1964)

62. Ellison, A.H., Zisman, W.A., 1. Phys. Chem. 59:1233 (1955)

63. Ellison, A.H., and Zisman, W.A., J. Phys. Chem. 60:416 (1956)

64. Ellison, A.H., and Zisman, W.A., J. Phys. Chem. 63:1J21 (1959)

65. Jarvis, N.L., and Zisman, W.A., J. Phys. Chem. 63:727 (1959)



68. Bernett, M.K., Jarvis. N.L., and Zisman, W.A., J. Phys. Chem. 66:328 (1962)

69. O'Rear, J.G., and Sniegoski, P.J., "New Partially Fluorinated Compounds," NRLReport 5795, July 18, 1962

70. Langmuir, L., J. Chem, Phys. 1:756 (1933)

71. Hedge, D.G., J. Colloid Sci. 12:417 (1957)


72. Phillips. N.J., and Rideal, E., Proc. Roy. Soc. (London) 232:149,159 (1955)

73. Jarvis, N.L., Fox, R.E., and Zisman, W.A , "Surface Activity at Organic Liquid-AirInterfaces," Advances in Chemistry, No. 43. Amer. Chem. Soc., Washington, D.C.,p. 317, 1964

74. Bowers, R.C., Jarvis, N.L., and Zisman, W.A., I. & EC Prod. Res. and Development

4:86 (1965)

75. McBain, J.W., and Hopkins, D.C., J. Phys. Chem. 29:188 (1925)

76. McBain, J.W., and Lee, W.B., Ind. Eng. Chem. 19:1005 (1927)

77. Hull, A.W., and Burger, E.E., Physics 5:384 (1934)

78. de Bruyne, N.A., Nature 180:262 (Aug. 10, 1957)

79. Zisman, W.A., "Constitutional Effects on Adhesion and Abhesion," in "Adhesion andCohesion," P. Weiss, editor, New York: Elsevier Publ, Co., pp. 176-208, 1962

80. Zisman, W.A., Ind. Eng. Chem. 55:19 (1963)

81. Sneddon, I.N., "The Distribution of Stress in Adhesive Joints," Ch. IX in "Adhesion,"D. Eley, editor, London: Oxford University Press, 1961

82. Mylonas, C., Proc. VII, Intern. Congr. Appl. Mech (London), 1948, Vol. 4

83. Mylonas, C., and de Bruyne, N.A., Ch. IV in "Adhesion and Adhesives," edited by N.A.de Bruyne and R. Houwink, Amsterdam: Elsevier Publ., Co., pp. 91-143, 1951

84. Griffith,. A.A., Phil. Trans. Roy. Soc. (London; A221:163 (1920)

85. Sharpe, L.H., and Schonhorn, H., "Surface Energetics, Adhesion, andAdhesive Joints,"Advances in Chemistry, No. 43, Amer. Chem. Soc., Washington, D.C., p. 189, 1964

86. Hendricks, J.O., Ind. Eng. Chem. 45:99 (1953)

87. Blake, G.B., Ahlbrecht, A.H., and Bryce, H.G., "Perfluoroalkyl Surface ActiveAgents for Hydrocarbon Systems," presented at 126th Meeting of the ACS, Divisionof Petroleum Chemistry, New York, N.Y., Sept. 1954

88. Talbot, E.L., J. Phys. Chem. 63:1666 (1959)

Security Classification

DOCUMENT CONTROL DATA- R&D*eatitp dl0.leclcotien Of title. body of ebstr.ct and Indesini nnotation must be ent.twd wh.n dhe o0e".41 eport is CMi.tle.d)

I OqIGINATIN G ACTIV1IY (Cooperate Author) 2, RCPORT SECuRITY €C L.ASsiPiCAiON

U.S. Naval Research Laboratory Uncl assi fied%ashington. D C. 20390 26 GROUP

3 REPORT TITLE

SURFACE CQIEMISTRY OF FLUOROCIIEMTCALS

4. DESCRIPTIVE NOTES (Type of teport and Inclusive dales)

An interim report on one phase of the problem.S AUTHOR(S) (Last notes. fir name. inlltil)

Jarvis, N.L. and Zisman, W.A.

6- REPORT OATE 78. TOTAL NO. OP PAGES 7b. NO. OP REP

October 21, 1965 36 88Ila. CONTRACT ON GRANT NO., 1. ORIGINATOR'iS REPORT NUMUER(S)

NRL Problem C02-10,PROJECT HO. NRL Report 6324RR 001-01-43-4151

C.~ti S tHER N EPORY NO(S) (Any ath.,nuamh.,. chat nway b,. .&ot~dn

10 AVA ILAUILITY/LIMITATION NOTICES

1I SUPPLEMENTARY NOTES 12 SPONSORING MILITARY ACTIVITY

Dept. of the Navy (Office of Naval Research)

I1 ABSTRACT

Research on the surface chemistry of fluorochemicals, including the related developme-t ofnew fluorine-containing compounds, polymers, and surface treatments, has progressed to the stagewhere a review of the subject is appropriate. A report has been prepared in the hope that it willserve not only as a record of progress, but will also clarify much of the fundamental informationthat is available, and will serve as a guide for future investigations and applications.

With the advent of the highly fluorinated organic compounds, it has been possible to preparesurfaces having extremely low free surface energies, much lower than had previously been possiblewith hydrocarbon derivatives. Because of their low surface energies, fluorinated solids have themost nonwettable and nonadhesive surfaces known. It has been proved that only the outermost mo-lecules of the surface of a solid need be highly fluorinated in order to achieve such low surfaceenergies. In fact the most nonwetting surfaces studied to date were those covered with an ad-sorbed monolayer of closely packed perfluoro acids, with the perfluoromethyl (-CF3 ) groups outer-most.

It has likewise been found that surface-active agents containing perfluorocarbon terminalgroups will lower the surface tension of water far below the values obtainable with hydrocarbon-type wetting agents. Certain fluorinated hydrocarbon derivatives have also been synthesized thatshow very high surface activity in a wide variety of organic liquids, the extent of surface ac-tivity being dependent upon the organophilic-organophobic balance in the fluorine-containingamphipathic molecule.

It is predicted that fluorine-containing munorers and fluorocarbon-substituted surface-activematerials will find increasing application in polymers and plastics, as additives for liquids, andas protective coatings on solids.

D D ¶I`'.. 1473 33Security Classification

I_ _ _ _ _ _

Security Classification14 LINK A LINK 9 LINK C

KEY WORDS 0.01 T ROL. E "' FiQiL.L W1i

Fluorine (comlpounds H

olo 1 yme I s

Surface firopert ie s0, gaiJ (comnpounds !Stir far'-'-,t iye vu{st-ances

Surfa(e tension I

INSTRUCTIONS

1. ORIGINATING ACTIVITY: Enter the name and address imposed by security classification, using standard statementsof the contractor, subcontractor. grantee, Department of Di' such as:fense activity or other organization (corporate author) issuing (1) "Qualified requesters may obtain copies of thisthe report. report from DDC. "

2a. REPORT SECTUITTY CLASSIFICATION: Enter the over- (2) "Foreign announce-ment and dissemination of thisall "vccurity classification o! the report. Indicate whether report by DDC is not authorized.""Restricted Data" is included. Marking is to be in accord-ance with appropriate security regulations. (3) "U. S. Government agencies may obtain copies of

this report dirictly from DDC. Other qualified DDC2h. GROUP: Automatic downgrading is specified in DoD Di- users shall request throughrective 5200. 10 and Armed Forces Industrial Manual. Enterthe groop number. Also, when applicable, show that optionalmarkings have been used for Group 3 and Group 4 as author- (4) "U. S. mulitary agencies may obtain copies of thisized. report directly from DDC. Other qualified users

3. REPORT TITLE: Enter the complete report title in all shall request throughcapital letters. Titles in all cases should be unclassified.If a meaningful title cannot be selected without classifica-tion, show title classification in all capitals in parenthesis (5) "All distribution of this report is controlled. Qual-immediately following the title. ified DDC users shall request through

4. DESCRIPTIVE NOTES: If appropriate, enter the type of --__"

report, e.g., interim, progress, summary, annual, or final. If the report has been furnished to the Office of TechnicalGive the inclusive dates when a specific reporting period is Services, Department of Commerce, for sale to the public, mdi-covered. cate this fact and enter the price, if know,.

S. AUTHOR(S); Enter the name(s) of author(s) as shown on It. SUPPLEMENTARY NOTES: Use for additional explana-or in the report. Enter last name, first name, middle initial. tory notes.If military, show rank and bratich of service. The name ofthe principal author is an absolute minimum requirement. 12. SPONSORING MILITARY ACTIVITY: Enter the name of

the departmental project office or laboratory sponsoring (pay-6. REPORT DATE. Enter the date of the report as day, inm for) the research and development. Include address.month, year; or month, year. If more than one date appearson the report. use date of publication. 13. ABSTRACT: Enter an abstract giving a brief and factual

summary of the document indicative of the report, even though7m. TOTAL NUMBER OF PAGES: The total page count it may also appear elsewhere in the body of the technical re-

should follow normal pagination procedures, Le., enter the port. If additional space is required, a continuation sheet shallnumber of pages containing information, be attached.

7b. NUMBER OF REFERENCES Enter the tot.,t number of It is highly desirable that the abstract of classified reports

references cited im the report. be unclassified. Each paragraph of the abstract shall end with

Ba. CONTRACT OR GRANT NUMBER: If appropriate, enter an indication of the -'itary security clatsification of the in-the applicable number of the contract or grant under which formation in the paragraph, represented as (TS). (Si (C). or (U)

the report was written. There is no limitation on the length of the abstract. How-

Ab, 8c, & Sd. PROJECT NUMBER: Enter the appropriate ever. the suggested length is from IS0 to 225 wordsmilitary department identification, such as project number.subproject numberr system numbers, task number, etc. 14 KEY WORDS: Key words are technically meaningful terms

or short phrases that characterize a report 4tn may be used as9s. ORIGINATOR'S REPORT NUMBER(S): Enter the Offt- index entries for cataloging the report. Key words must becial report number by which the document will be identified selected so that no security classiftcation is required. Idetiti-and controlled by the originating activity. This number must fiers, such as equipment model designation, trade name, militarybe unique to this report. project code name, geographic location, may be used as key

9b. OTHER REPORT NUMBER(S): if the report has been words but will be followed by an indication of technical con-

assigned any other report numbers (either by the originator text. The assignment of links, roles, and weights is optional.

or by the sponsor), also enter this number(s).

10. AVALLABILITY/LIMITATION NOTICES: Enter any lim-itations on further dissemination of the report, other than those

34Security Classification

surface chemistry of fluorochemicals

Documents